Résumé
It is now well supported that 20% of human cancers have an infectious causation (i.e., oncogenic agents). Accumulating evidence suggests that aside from this direct role, other infectious agents may also indirectly affect cancer epidemiology through interactions with the oncogenic agents within the wider infection community. Here, we address this hypothesis via analysis of large-scale global data to identify associations between human cancer incidence and assemblages of neglected infectious agents. We focus on a gradient of three widely-distributed cancers with an infectious cause: bladder (~2% of recorded cancer cases are due to Shistosoma haematobium), liver (~60% consecutive to Hepatitis B and C infection) and stomach (Helicobacter pylori is associated with ~70% of cases). We analyzed countries in tropical and temperate regions separately, and controlled for many confounding social and economic variables. First, we found that particular assemblages of bacteria are associated with bladder cancer incidences. Second, we observed a specific and robust association between helminths and liver cancer incidences in both biomes. Third, we show that certain assemblages of viruses may facilitate stomach cancer in tropical area, while others protect against its development in temperate countries. Finally, we discuss the implications of our results in terms of cancer prevention and highlight the necessity to consider neglected diseases, especially in tropics, to adapt public health strategies against infectious diseases and cancer.